Multifunctional DNA Polymerase-Lyase-Ligase For
Total Page:16
File Type:pdf, Size:1020Kb
www.nature.com/scientificreports OPEN Phaeocystis globosa Virus DNA Polymerase X: a “Swiss Army knife”, Multifunctional DNA Received: 4 May 2017 Accepted: 27 June 2017 polymerase-lyase-ligase for Base Published: xx xx xxxx Excision Repair José L. Fernández-García1, Ana de Ory1, Corina P. D. Brussaard2 & Miguel de Vega 1 Phaeocystis globosa virus 16T is a giant virus that belongs to the so-called nucleo-cytoplasmic large DNA virus (NCLDV) group. Its linear dsDNA genome contains an almost full complement of genes required to participate in viral base excision repair (BER). Among them is a gene coding for a bimodular protein consisting of an N-terminal Polβ-like core fused to a C-terminal domain (PgVPolX), which shows homology with NAD+-dependent DNA ligases. Analysis of the biochemical features of the purifed enzyme revealed that PgVPolX is a multifunctional protein that could act as a “Swiss army knife” enzyme during BER since it is endowed with: 1) a template-directed DNA polymerization activity, preferentially acting on DNA structures containing gaps; 2) 5′-deoxyribose-5-phosphate (dRP) and abasic (AP) site lyase activities; and 3) an NAD+-dependent DNA ligase activity. We show how the three activities act in concert to efciently repair BER intermediates, leading us to suggest that PgVPolX may constitute, together with the viral AP-endonuclease, a BER pathway. This is the frst time that this type of protein fusion has been demonstrated to be functional. DNA base lesions are the most common type of genomic damage and pose a challenge to genome stability. Lesions can arise afer exposure of DNA to genotoxicants [e.g., radiation, alkylating mutagens, by-products of cellular metabolism such as reactive oxygen species (ROS)], or can arise spontaneously under physiological conditions, such as the hydrolysis of the glycosidic bonds that leaves uninformative apurinic/apyrimidinic (AP) sites, and deaminations that lead to miscoding bases1, 2. Te base excision repair (BER) pathway is a multi-step enzymatic pathway that is primarily responsible for repairing a broad spectrum of non-bulky and non-helix-distorting DNA lesions produced by the oxidation, alkylation, deamination or hydroxylation of DNA bases3. Te general pathway consists of fve steps: 1) recognition and hydrolytic cleavage of altered base-sugar bonds by DNA N-glycosylases4; 2) recognition of the resulting AP sites by an AP-endonuclease, which cleaves at the 5′-side of the AP site to ren- der a 1-nt gap fanked by 3′-OH and 5′-dRP termini; 3) removal of the 5′-dRP moiety by a 5′-dRP lyase, leaving a ligatable 5′-P end; 4) short gap flling by a specifc DNA polymerase to restore the original (nondamaged) nucle- otide; and 5) fnal nick sealing by a DNA ligase (reviewed in ref. 3). Family X DNA polymerases (PolXs), including mammalian Pol β5 and Pol λ6, 7, bacterial PolXs8–10 as well as the African Swine Fever Virus (ASFV) PolX11, 12, are involved in the flling-in step of BER. Tese polymerases are relatively small, monomeric proteins that catalyze the insertion of a few nucleotides and lack an intrinsic proofreading activity13. Teir optimized architecture allows them to efciently accomplish the flling-in of short gaps that arise in BER. In general, PolX members present a common Polβ-like core14 with an N-terminal 8-kDa domain that specifcally recognizes the 5′-phosphate group of the gap, which allows the correct positioning of the enzyme on the gapped or nicked structure15–17. Te 8-kDa domain of several eukaryotic PolXs including Pol β and Pol λ18, 19 and yeast Pol420 and Trf421, is also endowed with a 5′-dRP lyase activity that removes the 5′-dRP 1Centro de Biología Molecular Severo Ochoa (CSIC-UAM), C/Nicolás Cabrera 1, Campus Cantoblanco, 28049, Madrid, Spain. 2Department of Marine Microbiology and Biogeochemistry, NIOZ Royal Netherlands Institute for Sea Research, Utrecht University, NL-1790 AB, Den Burg (Texel), The Netherlands. Correspondence and requests for materials should be addressed to M.d.V. (email: [email protected]) SCIENTIFIC REPORTS | 7: 6907 | DOI:10.1038/s41598-017-07378-3 1 www.nature.com/scientificreports/ group during short-patch BER6, 22. Te C-terminal polymerization domain of the Polβ-like core possesses the three universally conserved subdomains, fngers, palm and thumb, which are responsible for the binding and further elongation of the 3′ terminus of the upstream primer strand14. Te only known exception to this structural generality is the ASFV PolX, which lacks the 8-kDa domain and the fngers subdomain11. In several PolXs, the domain responsible for the polymerization reaction is fused to catalytic and/or protein-protein interaction domains. Tus, the Polβ-like core of the eukaryotic DNA polymerases λ, μ, termi- nal deoxynucleotidyl transferase (TdT) and yeast Pol IV, is fused to an N-terminal BRAC1 carboxy terminus (BRCT) domain that interacts with nonhomologous end joining-DNA bound factors to recruit these polymerases to double-strand breaks23–28. In Bacterial and Archaeal PolXs, the C-terminus of the Polβ-like core is fused to a polymerase and histidinol phosphatase (PHP) domain29, which contains 3′-5′ exonuclease, 3′-phosphatase and 3′-phosphodiesterase activities, and endows the polymerase with the capacity to process damaged 3′ termini10, 30–32. It also has an AP-endonuclease activity, which enables the enzyme to process an AP site and restore the original nucleotide9, 10. Aside from bacteria, archaea and eukaryotes, some viruses contain PolXs. Te nucleo-cytoplasmic large DNA virus (NCLDV) group constitutes a monophyletic group of viruses that infect a wide range of eukaryotes, such as unicellular marine protists, insects, fsh and mammals. Tey possess dsDNA genomes ranging from 100 kb to 2.5 Mb that, unlike smaller viruses, usually have many genes that encode for putative DNA repair proteins33, 34. Te genome of the viral strain 16T that infects Phaeocystis globosa (strain PgV-16T), a high-biomass-forming phytoplankton species with a central role in oceanic carbon and sulfur cycles35, has been recently sequenced36. PgV-16T is a giant virus from the Megaviridae family36 that belongs to the NCLDV group and has a linear dsDNA genome 459,984 bp in length. Its genome contains an almost full complement of genes (Xth/APE1-like AP-endonuclease, a hypothetical PolX and a NAD+-dependent DNA ligase) required to execute a potential viral BER pathway33, 34, 36. NAD+-dependent ligases are present in a minority of NCLDVs and most likely were acquired from a bacteriophage at the early stages of evolution of eukaryotes33, 37, whereas PolX could be acquired from diferent cellular organisms including eukaryotes33, 38. Interestingly, whereas PolX and ligase are encoded by two independent genes in the Marseilleviridae, Mimiviridae and Poxviridae families, which also belong to the NCLDV group, PgV-16T and its close relative CeV, a virus infecting the unicellular marine phytoplankton Haptolina (formerly Chrysochromulina) ericina, represent the frst examples where the DNA ligase gene is fused to the C-terminus of the PolX coding region36, 39. To expand the current knowledge on the role of PolXs in BER, we have characterized the putative PolX from PgV-16T (PgVPolX). Our results indicate that, in addition to the general polymerization properties shared by most PolXs, PgVPolX has an intrinsic 5′-dRP lyase and NAD+-dependent DNA ligase activity. Te coordina- tion of these activities enables the enzyme to perform the last three steps of BER: removal of the 5′-dRP moiety, flling-in of the resulting gap, and fnal sealing of the nick to restore the original genomic information. To the best of our knowledge, this is the frst characterization of this type of fused protein. Results P. globosa virus 401 gene codes for a family X DNA polymerase. Te complete genome sequence of the P. globosa virus revealed an ORF corresponding to gene 401, which would encode for a 1093 amino acid protein whose N-terminal 379 amino acids have homology with family X DNA polymerases (30% identity with human Pol β) (Fig. 1a); and whose C-terminal domain (475–1050) has homology with the NAD+-dependent DNA ligases36 (Fig. 1b). To determine whether this type of fused viral gene is functional, we cloned the entire ORF into the pET16(a)+ expression vector which carries an N-terminal (His)10-tag. Te recombinant protein was expressed in Escherichia coli BL21(DE3) cells and purifed as described in Materials and Methods. To examine the presence of a polymerization activity in PgVPolX, the purifed protein was incubated with a 5′-Cy5-labeled primer/template DNA (depicted in Fig. 2, top), 100 µM dNTPs and Mg2+ as a metal activator. As shown in Fig. 2a (lef panel), PgVPolX catalyzed the 5′-3′ extension of the primer molecule, demonstrating a distributive polymerization activity as it yielded polymerization products whose length was strongly dependent on the enzyme:DNA ratio. Tis result suggests that PgVPolX is well suited to accomplish short-stretch DNA synthesis in vivo. To determine whether the polymerization activity was inherent to PgVPolX, the purifed pro- tein was sedimented through a glycerol gradient, and the collected fractions were individually assayed for DNA polymerase activity on the same substrate. A single activity peak cosedimented with the mass peak of the purifed PgVPolX (Fig. 2b). Additionally, substitution of the two predicted metal ligands Asp245 and Asp247 with alanine (Fig. 1a) rendered a virtually inactive protein (Fig. 2a, right panel), and the residual polymerization activity was intrinsic to the mutant enzyme (Supplementary Fig. S1A). Overall, the results allow us to unambiguously assign a DNA polymerization activity to PgVPolX and rule out the presence of a contaminant DNA polymerase from E. coli. As expected for a PolX family DNA polymerase, the presence of a downstream oligonucleotide in 5-nt gapped structures promoted the rapid appearance of a band corresponding to the fully repaired gap (Fig.